This is a new one for me. There were a slew of updates to dom0 and somewhere along the line, my setup started having slow upload speeds to my internal network. Both sys-net and internal host are on the same wifi net, nothing in between.
sys-net → internal host = 33mb/sec
AppVm → sys-net → internal host = 3mb/sec
And this was after I changed the sys-net to a fedora 40 template (was debian 12) and also using an older kernel (6.6.77). With debian 12 and the latest kernel (6.12.18-1) transfer speed would be around 500kb and would stall often with these types of errors from strace
read(3, “\34\220:\311V\211/E,r\304\325\2\30\237\276\4s6G\334\271\340\352"\311\244\347\324.\244\215”…, 261120) = 261120
writev(4, [{iov_base=“\0\3\374\31”, iov_len=4}, {iov_base=“\6\0\0\0001\0\0\0\4\0\0\0\0\0\0\0\0\0\257P\0\0\3\374\0\34\220:\311V\211/”…, iov_len=261145}], 2) = 73088
— SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} —
alarm(1) = 0
rt_sigreturn({mask=}) = 73088
getpgrp() = 1726
ioctl(1, TIOCGPGRP, [1726]) = 0
tempo 0% 255KB 167.0KB/s 29:54 ETA) = 190
writev(4, [{iov_base=“de\303 \224\310\262k\367\37\f\3350w\267\331\344\271]\22\375\331@\376s\202\f\3427\263\377\31”…, iov_len=188061}], 1) = ? ERESTARTSYS (To be restarted if SA_RESTART is set)
— SIGALRM {si_signo=SIGALRM, si_code=SI_KERNEL} —
alarm(1) = 0
rt_sigreturn({mask=}) = -1 EINTR (Interrupted system call)
getpgrp() = 1726
ioctl(1, TIOCGPGRP, [1726]) = 0
tempo 0% 255KB 150.3KB/s 33:13 ETA) = 190
Prior to the slew of changes to dom0, I was able to get better speeds from the AppVm to the internal host used to get great speeds. Maybe something in the latest updates of dom0 causes this? I can help in debugging if required. Just need to know what to give you guys as info to assist.
Edit: Qubes OS version 4.2.4